How it works
No step waits for a consultant. The system guides you from raw data to traceable recommendations.
Step 1
01Upload data
Upload transaction data as CSV, text or Excel using standard exports. Files are stored encrypted in Microsoft Azure in Sweden, within the EU.
Step 2
02Confirm columns
The system suggests what each column means. You review and confirm the mapping.
Step 3
03Quality control
Data is checked automatically. Serious errors stop the run; smaller deviations are documented.
Step 4
04Deterministic analysis
The engine calculates findings and potential. Every figure is traceable.
Step 5
05AI interviews
Send personal links to key people. Intelligent chatbots gather business context while analysis runs. Unlike a scheduled meeting, respondents can pause at any point and continue later using the same link — no diaries to coordinate.
Step 6
06Report and results chat
Findings and interviews become a report, dashboard, presentation and workshop pack. An AI chatbot explains results and discusses initiatives.
What you need
- Transaction data at order-line or invoice level.
- Someone who can confirm what the columns mean.
- Optional: people who can answer AI interviews about the business.
Fixed product, fixed price
The standard analysis costs 899 GBP excluding tax. A workshop pack is included. A consultant-led workshop or executive presentation can be added.
Is your data good enough?
Your data is rarely the barrier — it is almost always sufficient. What you need is a raw, line-level export from your ERP system. In short:
Real names are not required
If customer or product names are sensitive, you can mask them — replace them with your own codes such as "Customer A" or "Product 12". The only requirement is consistent masking: the same customer must always have the same masked name, throughout the file and in every sheet. Otherwise, the rows cannot be linked and the right conclusions cannot be drawn. Just tell us which columns are masked during the mapping step.
- One row per item sold — invoice or order lines, not a pivot table or summary.
- One full year — a calendar year or the rolling last 12 months — so a year of seasonality and pricing effects is visible.
- Each row shows when, to whom, which product, how many and at what price.
- Unit cost if available — without it we show price but not margin.
- Raw data straight from the system is better than processed files. Columns can have any name — we help you map them.
- Excel or CSV, up to 100 MB per file. Multiple sheets are fine. Larger datasets can be split into several files, e.g. one year per file, and merged with Extend history.
Show the detailed data request
Feel free to send the request to your controller or system administrator — it specifies exactly how the file should look.
DATA REQUIRED FOR THE PRICING ANALYSIS File format: Excel (.xlsx/.xls) or CSV; CSV is recommended above 100,000 rows. One row = one order/invoice line (item position), not an entire order. Columns needed (column names can be in any language — you confirm which column is which yourself): • Date (transaction date, ideally including year and month) • Customer (customer ID or name) • Product/service (item ID or name) • Quantity • Revenue per line (amount before/after discount, as invoiced) • Unit price or list price, if available • Cost per line or purchase price (optional, but enables margin analyses) • Discount per line (optional, if discounts are not already in revenue) Currency: all amounts in one currency, or a column with the currency code. Coverage: ideally at least 12 months of completed transactions; multiple years enable better comparisons. Larger groups: choose one coherent business unit or market at a time, rather than the entire group in a single file. Do not include personal data. Remove or replace names of individuals in the columns you upload.